| Summary: | 碩士 === 輔仁大學 === 統計資訊學系應用統計碩士在職專班 === 103 === Background:
Mortality of peptic ulcer bleeding majorly results from non-bleeding causes. In cirrhotic patients, it is evident for infection associated with mortality of upper gastrointestinal hemorrhage, including peptic ulcers. However, there was lack of study for the association between infection and mortality risk of peptic ulcer bleeding in non-cirrhotic patients. Literature showed upper gastrointestinal endoscopy and acid-suppressive agent universally administered for peptic ulcer bleeding could be related to pneumonia.
Aim:
Using pneumonia as the representative of infectious disease, this research aimed to study the association between pneumonia and risk of mortality in non-cirrhotic patients with peptic ulcer bleeding and find high-risk patient groups of pneumonia.
Method:
Based on Taiwan national healthcare insurance database, we retrospectively selected hospitalized patients with primary diagnosis of non-perforated peptic ulcer bleeding between 2011 and 2012. Adult non-cirrhotic patients were statistically analyzed by logistic regression analysis and classification and regression tree method (CART).
Result:
There were total 5690 non-cirrhotic admitted patients with 78 in-hospital mortalities (1.4%). The explanatory variables include age, gender, ulcer type, admission days, stroke, coronary artery disease, chronic lung disease, pneumonia, renal insufficiency, endoscopic hemostasis therapy, surgery for ulcer hemorrhage, and ventilator use. Survival control group composed of randomly selected non-matched 780 cases. Logistic regression showed predictors of in-hospital mortality as following: pneumonia (odds ratio 36.28), admission days 14-day or more (odds ratio 6.85) and renal insufficiency (odds ratio 4.05). CART showed six decision factors for risk of pneumonia: age, gender, stroke, coronary artery disease, renal insufficiency, and endoscopic hemostasis therapy.
Conclusion:
For non-cirrhotic patient hospitalization for peptic ulcer bleeding, pneumonia, admission length 14-day or more and renal insufficiency are predictors of in-hospital mortality. High-risk patients with concurrent peptic ulcer bleeding and pneumonia were as following: 1. Age 55 to below 70, male, renal insufficiency or stroke; 2. Age 70 or more, endoscopic hemostasis therapy or coronary artery disease.
|
|---|
